1

According to most voted ansewer here: Is the literal 0xffffffff int or unsigned in C++. There says "Assuming your implementation has 32-bit int, since 0xffffffff does not fit in an int, its type should be unsigned int." unfortunately, can't comment there due to new here.

Question 1: but why does not 0xffffffff fit in an int? Because its value is 4294967295? If yes, then we are interpreting it as unsigned already. why don't we interpret it as signed -1?

I also run the following code, (compiler is clang, 32bit int): 
cout << std::is_signed<decltype(0xffffffff)>::value << " " << 0xffffffff << endl;
cout << std::is_signed<decltype(0xffffffe)>::value << " " << 0xfffffffe << endl;

bool same = std::is_same_v<decltype(0xffffffff), decltype(0xfffffffe)>;
std::cout << "0xFFFFFFFF and 0xFFFFFFFE are the same type? " << std::boolalpha << same << '\n';

decltype(0xffffffff) a = -1;
decltype(0xfffffffe) b = -1;
cout << a << " " << b << endl;

The output is:

0 4294967295
1 4294967294
0xFFFFFFFF and 0xFFFFFFFE are the same type? true
4294967295 4294967295

Question 2: why do 0xffffffff and 0xfffffffe have different signs but are the same type?

Question 3: if the sign of 0xfffffffe is 1, then why isn't itself a negative number?

Question 4: if the sign of 0xfffffffe is 1, then why isn't b -1?

S55
  • 53
  • 5
  • C++ does not interpret negative numbers. It only interprets positive numbers, and then afterward is applies the *unary negation* to the positive number. So `-1` is interpreted as *unary negation* applied to a *positive* `1`. (This does have a few unfortunate edge-case consequences that need to be worked around, for example, look at how INT_MIN is implemented in your compiler's header file.) – Eljay Jun 13 '21 at 15:46
  • 1
    Also `0xffffffe` is not `0xfffffffe`. – Eljay Jun 13 '21 at 15:47
  • Q2: They are not of different signs, they are of different types. `int` vs `unsigned int` – tobias Jun 13 '21 at 15:49
  • On the second line, `std::is_signed`, you left out one digit. `0xffffffe` (7 digits) is indeed a positive `signed int`. Maybe you should fix this and then see if you still have a question. – Nate Eldredge Jun 13 '21 at 15:56
  • You may also want to read about [integer literals](https://en.cppreference.com/w/cpp/language/integer_literal) and how they get their associated "best fit" type. (It's a bit convoluted.) – Eljay Jun 13 '21 at 15:58

0 Answers0